Zebrafish offer a powerful model system in which it is possible to conduct large-scale phenotype-driven screening of recessive mutations. We have undertaken a chemical mutagenesis screen using N-nitroso-N-ethylurea (ENU) to uncover mutants in blood formation. Adult male fish were exposed to ENU to mutagenize spermatagonia and then crossed to wild-type females to generate F1 fish. Haploid embryos from F1 females were screened at 27 hours of development by RNA in situ hybridization using the myeloid-specific marker, leukocyte-specific plastin (l-plastin), and core binding factor b (cbfb). We have screened about 510 F1 female embryos and a total of 47 putative mutants have been identified. One of the mutant lines, L32, is bloodless and may have a defect at or above the level of hematopoietic stem cells. We have mapped the mutation to a region of less than 0.08 cM which corresponds to a distance of about 60 kb. We have also undertaken an effort to generate a mutant library archived as a sperm bank and a DNA bank from ENU-mutagenized F1 males, in order to screen for mutations in specific genes of interest. DNA will be used to screen for mutations by sequencing or other heteroduplex screening methods. The corresponding sperm samples will be used to revive the mutant fish by in vitro fertilization. To date, we have generated archived DNA, sperm and remaining body tissues from 1235 F1 mutants in 96-well format. We are currently in the process of sequencing these DNA for mutations in genes involved in blood and cancer. We are also in the process of generating additional mutants to increase the size of the library, thereby giving us a better chance to identify mutations in any given gene. Finally, we are generating transgenic fish expressing green fluorescent protein gene and leukemia fusion genes under the control of cbfb promoter, with targeted BAC constructs. Such transgenic fish will be useful tools for our genetic screens and for the generation of leukemia models.